The precise assembly of neural circuits provides the basis for nervous system function and animal behavior. Laminar arrangement of neural connections is a primary strategy for organizing neural circuits in vertebrates and invertebrates. Previous research has illuminated how particular neuron types target to and arborize within specific layers in isolated contexts. However, how the targeting and morphogenesis of different neuron types is coordinated to establish layered networks of connections is unknown. Addressing this gap in knowledge is fundamentally important to understanding how neural circuits are established. The goal of our research is to identify general molecular and cellular principles underlying the construction of layered neural networks. To accomplish this, our strategy is to determine how cells are coordinated to specific layers, and identify commonalities in how different layers assemble to illuminate general mechanisms. This approach requires precise knowledge of the cell types that innervate specific layers and genetic access to these cell types during development. Therefore, we study layer assembly in the Drosophila visual system, wherein well-characterized genetically accessible cell types synapse within specific layers in a stereotyped manner. In the Drosophila medulla, more than 60 uniquely identifiable neuron types synapse within 10 parallel layers. Previous studies indicate that medulla layers are refined during development from broad domains through a precise sequence of interactions between specific cell types. Similar findings in the mouse retina suggests this is a conserved developmental strategy for building synaptic layers. The main thrust of the proposal is to determine the molecular logic governing broad domain organization and the refinement of layers from these regions. We recently showed that Drosophila Fezf (dFezf), a conserved transcription factor, controls the assembly of a specific layer by coordinating the layer-specific innervation of different cell types. Based on preliminary findings, we hypothesize that (1) dFezf acts through a network of transcriptional regulators to control a gene program that regulates early and late stages of layer refinement, and (2) the use of transcriptional modules (like dFezf) to coordinate layer-specific innervation represents a general mechanism for constructing discrete layers. We will test this in 3 Specific AIMs. In AIMs I and II, we use dFezf as a handle to address the molecular underpinnings of (I) broad domain organization within the early medulla, and (II) the stepwise refinement of a specific layer. In AIM III we determine if transcriptional modules analogous to dFezf function generally to orchestrate the assembly of medulla layers. As the Drosophila visual system is analogous to the vertebrate retina in structure and function, and research in the mouse cortex is consistent with Fezf2 regulating the assembly of laminar circuitry, we expect our findings will have broad significance for the development of diverse nervous systems. In the long-term, we expect our findings to inform strategies for re-wiring neural circuits to restore brain function in the context of disease.